I have made the mistake of trying to get my head around the hideously complicated development path of the North American training series – there are maybe 100 variants, each with an alphabet soup of North American model number, charge number and military designation.
What I am trying to establish is when the six-inch extension made by moving the rudder post back happened. This was a subtly brilliant idea for reasons which are not immediately obvious, and changed the design from something of a liability with horrible (often deadly) spinning characteristics into an aircraft that could be spun to train recovery safely. It is often said that the success of the T-6 was down to this.
The thing is it was the British who led on the research that made this the way forward with the design – and it wasn’t just about tail length, there was something a lot more subtle in the design change that made the difference, to do with positive and negative damping and ‘masked’ rudder areas.
Sources are both vague and contradictory about how, where and when the change happened in US production. I know that the BC-1A, BC-2 and BT-14 had the new tail. It was also probably tried on the fabric-fuselaged BT-9D. There are confusions and complications around re-allocations of production runs, modifications mid-bloc, missing model numbers and charge number in sequences, and even 1940 fiscal year serial numbers on aircraft apparently ordered in 1937. The records are, frankly, awful.
My theory is that somewhere between February and September 1939 a hasty mod was made, in order to satisfy the British (who had lost a number of Harvard I’s in spinning accidents). The haste meant that all kinds of anomolies appeared in production and model number records, as it affected the various strands of production for the US and China as well (BT-series, BC-series, SNJ series). The British then went on to order 600 Harvard II’s in November – which was always the prize for getting it right.
It is only a theory – does anyone have any primary source information on this small but genuinely crucial modification – I have read the standard texts and they all just add to the fog of model numbers and conflicting timelines.
All help gratefully received and acknowledged in the thing I’m working on.
By: Dave Hadfield - 28th January 2018 at 17:25
Moving the rudder farther back also helps on the ground, reducing the likelihood of ground-loops, which the Harvard is known for.
This was one of the reasons it was done on the P-40 as well.
By: Beermat - 28th January 2018 at 17:22
Boscombe/Farnborough were aware of the problem before Feb 1939 – the crash that killed Peter Alston (who was an aerodynamicist) was during a spinning trial intended to investigate the problem. As for the turnaround time – essentially the same team designed and built the P-51 from scratch in 102 days.
There is no specific evidence about the tail either way – there is no evidence of the Americans doing anything at all, this new design just seemed to ‘appear’ if one goes by the records that have made it into secondary sources. There are papers coming out of the RAE, one in 1937 and one in 1940, that for improved spin recovery prescribe less fin/rudder area above the elevator (at least as of the 1940 report) and more unmasked rudder area (defined by a diagonal up from the trailing edge of the fully-depressed elevator) which the triangular shape and a fin further back relative to the horizontal surfaces gave – https://www.aerosociety.com/media/4839/on-the-early-history-of-spinning-and-spin-research-in-the-uk-part-2.pdf
[ATTACH=CONFIG]258607[/ATTACH]
(The NACA came up with a report echoing these shortly after).
Hence my looking for the missing discussions and proposals specific to the Harvard II/BT-14/BC2, but they do remain elusive. While they do I am not ruling out anything, including the influence of the work being done in the UK.
By: Graham Boak - 28th January 2018 at 14:48
I still think that there isn’t enough time for Boscombe/Farnborough to become aware how serious the problem was, initiate a set of trials, come to a conclusion, and get it all back to the US. Neither do I believe that the 6-in extension is specifically to cure the spin rather than more generally improve longitudinal/directional stability. Not that it would hurt, but I think that fixing the stall (ie the wing changes) would be more to the point. OK, I was a performance man not in stability and control. You say there’s no direct evidence that the long fuselage came from the US – there’s even less that it came from the UK. Everyone concerned was looking at the same problem – when it’s railroad time everyone railroads. It’s not as though the solution was unusual – it’s possible that the British banged the table a bit rather than coming out with anything quite so specific but where’s the evidence for even that?
By: Beermat - 28th January 2018 at 12:44
If I understand correctly, both the NA-44 and the SNJ-1 did have a metal fuselage but did NOT have the six-inch movement aft of the rudder post (and the triangular tail) in question? Please see the shot from Hagedorn, below.
I still cannot find anything demonstrating that this particular solution – the specific change to the tail – was in hand previous to February 1939. The BC-2 wasn’t ordered until October 1939.
I agree that on first reading of Hagedorn everything seems to imply earlier development, but there isn’t anything specific to support it.
[ATTACH=CONFIG]258601[/ATTACH]
By: Graham Boak - 27th January 2018 at 22:52
I’ve had a look in the Warbird Tech – and couldn’t find the spin study I mentioned, but did find a nice summary of the various changes, and which production variants they applied to. My suspicions was that it was the USN that was driving the tail changes, because they were the first to fit the bigger R1340. This was on the NJ-1, and it’s near twin the NA-26 (the first with retractable undercarriage). The first with the longer metal fuselage was the NA-44, which also had the bigger R-1820 Cyclone engine and was aimed at overseas combat sales. Hagedorn doesn’t quote a date here, but the manufacturer’s number is 44-747, whereas the first NA-49 Harvard Mk.I was 49-748. US production variants with the long fuselage and new wing began with the SNJ-1 and the BC-2. The SNJ-1 contract was signed September 23rd 1938, the Manufacturer’s numbers following the first batch of Harvard Mk.Is. The first Harvard Mk.I didn’t arrive until December 1938, the second in January 1939, with the first lost to a spin in February. So North American already had the solution before the British realised there was a problem, but the actual production was overlapping. That the SNJ-1 was first supports my feeling that it was the USN that pushed for change.
Arado, and Aermacchi later, pressed the argument that for safer spins the vertical fin should be further forward than the horizontal tailplane, but this isn’t what was going on in the case of the long-fuselage Texans. I see it as a counter to the additional forward side area of the bigger engine(s). The shape of the rudder, linked to the earlier intermediate solution of more area below the tailplane, does hint at earlier problems of rudder authority, but that interim rudder was already on the Harvard Mk.I.
By: Beermat - 27th January 2018 at 22:02
The wing changes were important, but the thing is the tests and the theory were documented in the US at the time. They only really fixed the problem in combination with the new tail – and the very odd thing is there appears to be no documentation of these changes or the reasons why they happened.
However on this side of the pond the RAE were working on spin departure, and calculated the effects of different orientations of the elevators and rudder. The new-shape Harvard had the rudder post six inches further back relative to the horizontal surfaces, and less rudder area above the elevators – both of would have made a significant difference, at least ‘on paper’, to the aircraft’s spin recovery times.
I became interested because, while there is no sign that North American were even working in this area (instead focusing, with justification, on the wing), the Harvard II tail geometry looks like a case-study on the British theory. When you add in the almost Hollywood angle that the RAE ‘Spin Tunnel’ was presided over by Gwen Alston, who lost her husband to a spinning Harvard I in February 1939, the story becomes even more interesting. However, the big ‘hole’ in the record – who designed the new tail and when – means it remains just one possibility – hence my search for the US evidence either way.
Thanks folks.
By: J Boyle - 27th January 2018 at 17:16
Great question.
I’m still waiting for North American Aircraft in the Putnam series.
It was announced (and included in a list of upcoming titles on the dust jacket of another book in the series) in the early 1990s, but never appeared.
Can the forum sleuths determine who wrote it and what became of the manuscript? Anyone have friends inside the publisher?
Certainly the firm that designed three of the great aircraft of WWII (the Mustang, Mitchell and Texan/Harvard), not to mention the Sabre and it’s developments, deserves to have a quality volume on its aircraft.
By: Beermat - 27th January 2018 at 16:56
Thanks Graham. I am on the road at the mo but will answer properly later. Good stuff!
By: Graham Boak - 27th January 2018 at 11:22
It’s certainly a complex story, partly because of concurrent developments and continued production of older variants alongside the newer ones. The British certainly were aware of the handling problems of the Harvard Mk.I and worked to alleviate them, but seem to have introduced nothing new and the changes to the Harvard line followed the developments in the US. It is possible to ignore most of the NA contract numbers and follow the main development lines – watching carefully for cross-overs! The design suffered spin problems from the start, and various changes were made as BT-9 production continued, including slats (as later adopted, not entirely satisfactorily, on the Harvard Mk.I) and washout on the outer wing panels. The move to the larger engine (BC-1 etc.) unsurprisingly seems to have accentuated these problems, as seen in the changes to the trailing edge of the rudder on the early SNJ and Harvard Mk.I. Stepping back to the small engined-line, minor changes to the wingtips were seen on the NA.64 Yale, and also I believe on its near twin the BT-14, both of which received the longer metal-covered rear fuselage and triangular rudder. (Quite why the BT-14 introduced the lower engine mounting is one of the smaller mysteries: sorry for the digression.)
However, the key change, following US wind tunnel tests, lay in reducing the sweepback of the wing (keeping the wash-out), thus moving the aerodynamic centre forwards to match the forward movement of the cg brought about by the heavier engine. When added to the greater tail moment arm of the longer rear fuselage, this seems to have effectively cured the aircraft’s problems.
As the best small single volume I recommend Dan Hagedorn’s Warbird Tech Vol.11, which itemises each individual contract number variant and includes evidence for the studies done for the wing redesign. It would be good to see each of the aerodynamic/engineering changes chronologically itemised and judged, but this kind of engineering approach to aircraft history doesn’t seem to have many fans amongst readers or writers.
Sign In
January 27, 2018 at 10:11 am